preheader BJH 1

High-dose methotrexate does not reduce CNS relapse in paediatric ALL

Results of the randomised phase III UKALL 2011 trial demonstrated that high-dose methotrexate does not improve central nervous system relapse but pre-specified analyses suggest it may improve bone marrow relapse for some subgroups of B-lineage patients when given following standard dexamethasone induction. Furthermore, the omission of pulses was non-inferior for bone marrow relapse.

The UKALL 2011 trial randomised children and young adults (up to 25 years old) with Acute Lymphoblastic Leukaemia (ALL) or Lymphoblastic Lymphoma (LBL). One of the goals of the study was to reduce steroid-related induction toxicity (R1: Short [14 days] vs. Standard [28 days] dexamethasone). R1 results were presented at ASH 2017 and the short course of dexamethasone treatment did not reduce toxicity. For the second and third aims of the study, researchers evaluated high-dose and standard methotrexate (HDM vs. standard interim maintenance) for reduction of central nervous system (CNS) relapse risk (R2IM) and determined whether dexamethasone pulses reduced maintenance morbidity (R2pulses). Results of R2IM and R2pulses were reported at ASH 2022.

Study design

Patients were stratified by National Cancer Institute (NCI) risk, cytogenetics and end of induction minimal residual disease (MRD) to three different intensity backbone regimens (A, B and C). Patients with MRD high risk (end of consolidation MRD >0.5%, N=14) were not eligible for randomisation and received off protocol therapy. R2 was a factorial randomisation stratified by factors including NCI and MRD risk groups and resulted in 4 arms: HDM with dexamethasone pulses, HDMTX without dexamethasone pulses, standard interim maintenance (SIM) with pulses (standard of care) and SIM without pulses. SIM was for 2 months with oral mercaptopurine/methotrexate (MTX), monthly pulses and single intrathecal MTX in regimens A and B, and 5 doses of escalating IV MTX (Capizzi) + vincristine + 2 doses of Pegylated asparaginase in Regimen C. All patients received intrathecal MTX in maintenance. In the HDMTX arms, MTX was given at a dose of 5g/m2 x 4 doses 2 weeks apart with low dose 6-MP and 2 doses of Pegylated asparaginase (Regimen C only). For the R2 pulses randomisation, researchers investigated the omission of monthly pulses of vincristine (1.5 mg/m2) and dexamethasone (5 days 6 mg/m2). The primary endpoint for R2IM was the CNS relapse rate (CNSR) and the primary endpoint for R2 pulses was the bone marrow relapse rate in patients with ALL. Researchers also evaluated event-free survival (EFS).


In total, 2750 patients were randomised to R1 (N= 1902) or R2 (N= 1570). Median age of the patients was 5 years and 83% had B-ALL. Median follow-up time was 76 months for R1 and 72 months for R2. Overall, researchers found no difference in CNSR (HR[95%CI]: 0.99[0.65-1.51], p= 0.97, 5-year rates: 5.6% vs. 5.6%), between the HDMTX and SIM arms. However, there was an interaction between the R1 and R2IM randomisations (p= 0.006 for EFS) with inferior outcomes for patients treated with short dexamethasone followed by HDMTX, particularly in those who did not receive pulses. If the analysis was limited to patients who received standard dexamethasone (N= 995, including those that received it after closure of R1) there was a significant reduction in BMR rate with HDMTX (HR[95%CI]: 0.62[0.40-0.95], p= 0.029). In addition, there were trends for improvements in EFS (HR[95%CI]: 0.75[0.53-1.04], p= 0.087)  and OS (HR[95%CI]: 0.61[0.37-1.03]), p= 0.067) but no difference in CNSR. Subgroup analyses revealed that the effect was stronger for B-lineage, NCI high-risk, and MRD low-risk patients. Overall, the BMR was non-inferior in R2pulses (+2.1% increase BMR at 5 years for ALL patients (95%CI: -1.4% to 4.7%), HR[95%CI]: 1.22[0.89-1.67], p= 0.22). When all four treatment arms were considered only in patients who received standard dexamethasone in R1, SIM without pulses has a lower EFS, whilst there is no appreciable benefit of pulses patients who received high-dose methotrexate; 5-year EFS difference: -2.8% (95% CI: -7.4% to 4.4%) and BMR (ALL): +0.4% (-6.7% to 4.3%).


HDMTX does not decrease CNS relapse but may reduce BM relapse for some subgroups of B-lineage patients following standard dexamethasone induction. Omission of pulses is not inferior for BM relapse. Omission of pulses slightly reduced EFS but the advantage of pulses is less evident in patients who received HDMTX and standard dexamethasone induction upfront.


Kirkwoord A, et al. High Dose Methotrexate Does Not Reduce the Risk of CNS Relapse in Children and Young Adults with Acute Lymphoblastic Leukaemia and Lymphoblastic Lymphoma. Results of the Randomised Phase III Study UKALL 2011. Presented at ASH 2022; Abstract 214.

Speaker Amy Kirkwood

Amy Kirkwood

Amy Kirkwood, MSc, University College London, London, United Kingdom


See: Keyslides


Back to Top